typical growth kinetics and operational conditions, namely a specific growth
rate of 0.236 h –1 and a superficial velocity of 0.0236 m s –1 , they predicted that
intervals between mixing of the order of one hour would lead to maximum bed
temperatures 2.4 °C higher than those expected for completely static operation.
This behavior occurs because, after a mixing event, the inlet air cools the bottom of the column back down to the air inlet temperature, but in doing so loses
its capacity to cool the upper regions of the column. After each mixing event a
cooling front travels up the bed, taking over 20 min to reach the top of a column
only 35 cm high, during which period the waste heat generation at the top of the
column has caused a significant increase in temperature [161]. The maximum
temperature achieved in the bed could only be decreased relative to that for
static operation if there were at least 10 mixing events per hour. At 30 mixing
events per hour the predictions were quite similar to those for a continuously
mixed bed [161].
5.5.3
Rocking Drum Bioreactor
The rocking drum bioreactor consists of three concentric cylinders lying
horizontally [107, 162–164] (see Fig. 5). The inner and middle cylinders are perforated, and are contained within the outer unperforated cylinder. The substrate
is loosely packed into the space between the inner and middle cylinders. Air
and small amounts of water are introduced into the innermost cylinder with the
air moving radially outwards through the substrate bed and the water percolating downwards due to gravity. When the air exits through the perforations
in the middle cylinder it flows in the space between this cylinder and the
outermost cylinder to the air outlet. The outer two cylinders rotate around the
stationary inner cylinder, causing a gentle mixing action within the bed. Typical
operation involves 3/4 turns backward and forth, hence the name of “rocking
drum.”
The rocking drum bioreactor has some similarities with a packed bed in that
the air flowing radially outwards is like the air flowing axially within a packed
bed. An advantage of the rocking drum reactor over the packed bed is that
water can be replenished reasonably evenly with a reasonably gentle mixing
action [107].
The work done to date with rocking drum bioreactors has addressed the
question of automatic control of SSF bioreactors. Barstow et al. [162] used a control scheme where the air flowrate was increased from 6.2 l min –1 to 8.5 l min –1
whenever the temperature exceeded the set point of 37 °C, and between 9 h and
43 h of fermentation the inlet air was dry. Ryoo et al. [163] pointed out that finer
control could be achieved by mixing dry and wet air to achieve the desired relative humidity at a constant overall air flowrate. They managed to maintain the
fermentation temperature within 0.5 °C of the set point. However, this work was
done in a bioreactor with a holding volume of only 1.3 l, and it is difficult to say
how well the bioreactor will operate on a large scale.
The modeling work done for the bioreactor assumes that the bed held between the innermost and middle perforated drums is well mixed, although no
116
D.A. Mitchell et al.
rate of 0.236 h –1 and a superficial velocity of 0.0236 m s –1 , they predicted that
intervals between mixing of the order of one hour would lead to maximum bed
temperatures 2.4 °C higher than those expected for completely static operation.
This behavior occurs because, after a mixing event, the inlet air cools the bottom of the column back down to the air inlet temperature, but in doing so loses
its capacity to cool the upper regions of the column. After each mixing event a
cooling front travels up the bed, taking over 20 min to reach the top of a column
only 35 cm high, during which period the waste heat generation at the top of the
column has caused a significant increase in temperature [161]. The maximum
temperature achieved in the bed could only be decreased relative to that for
static operation if there were at least 10 mixing events per hour. At 30 mixing
events per hour the predictions were quite similar to those for a continuously
mixed bed [161].
5.5.3
Rocking Drum Bioreactor
The rocking drum bioreactor consists of three concentric cylinders lying
horizontally [107, 162–164] (see Fig. 5). The inner and middle cylinders are perforated, and are contained within the outer unperforated cylinder. The substrate
is loosely packed into the space between the inner and middle cylinders. Air
and small amounts of water are introduced into the innermost cylinder with the
air moving radially outwards through the substrate bed and the water percolating downwards due to gravity. When the air exits through the perforations
in the middle cylinder it flows in the space between this cylinder and the
outermost cylinder to the air outlet. The outer two cylinders rotate around the
stationary inner cylinder, causing a gentle mixing action within the bed. Typical
operation involves 3/4 turns backward and forth, hence the name of “rocking
drum.”
The rocking drum bioreactor has some similarities with a packed bed in that
the air flowing radially outwards is like the air flowing axially within a packed
bed. An advantage of the rocking drum reactor over the packed bed is that
water can be replenished reasonably evenly with a reasonably gentle mixing
action [107].
The work done to date with rocking drum bioreactors has addressed the
question of automatic control of SSF bioreactors. Barstow et al. [162] used a control scheme where the air flowrate was increased from 6.2 l min –1 to 8.5 l min –1
whenever the temperature exceeded the set point of 37 °C, and between 9 h and
43 h of fermentation the inlet air was dry. Ryoo et al. [163] pointed out that finer
control could be achieved by mixing dry and wet air to achieve the desired relative humidity at a constant overall air flowrate. They managed to maintain the
fermentation temperature within 0.5 °C of the set point. However, this work was
done in a bioreactor with a holding volume of only 1.3 l, and it is difficult to say
how well the bioreactor will operate on a large scale.
The modeling work done for the bioreactor assumes that the bed held between the innermost and middle perforated drums is well mixed, although no
116
D.A. Mitchell et al.